Pedestrian injuries are among the leading causes of pediatric injury and mortality in American children ages 7-8, but existing behavior-oriented interventions achieve only modest success. One limitation to existing interventions is that they fail to provide children with the repeated practice needed to develop the complex perceptual and cognitive skills required for safe pedestrian activity. Virtual reality (VR) offers a highly promising technique to train children in pedestrian safety skills. VR permits repeated unsupervised practice without risk of injury; automated feedback to children on success or failure in crossings; adjustment of traffic density and speed to match children's skill level; and an appealing and fun environment for training. The proposed research is designed to test the efficacy of virtual reality as a tool to train child pedestrians in safe street-crossing behavior. A randomized controlled trial will be conducted with four equal-sized groups of children ages 7-8 (total N = 240). One group will receive training in an interactive and immersive virtual pedestrian environment. The virtual environment, already developed, has been demonstrated to have face, construct, and convergent validity. The second group will receive pedestrian safety training via video and computer strategies that are most widely used in American schools today. The third group will receive what is judged to be the most efficacious treatment currently available, individualized behavioral training at street side locations. The fourth and final group will serve as a no-contact control group. All participants in all groups will be exposed to a range of field- and laboratory-based measures of pedestrian skill during baseline and post-intervention visits, as well as during a six-month follow-up assessment. Primary analyses will be conducted through linear mixed models designed to test change over time in the four intervention groups. PUBLIC HEALTH RELEVANCE: Pedestrian injuries are among the leading cause of pediatric mortality for American children ages 7-8. This project will study the efficacy of virtual reality as a means to train children in safe pedestrian behavior. Results will have significant implications for child pedestrian injury prevention.